A recent study suggests that electronic systems cause just as many errors as paper-based ones, but deeper analysis doesn't support that conclusion.
David Pogue, the technology editor for The New York Times, likes to talk about consumers' pain points, including the hassle of setting up a home network and e-mailing a video to a friend. Healthcare IT executives have their own long list of pain points, and the latest study published online in the Journal of the Medical Informatics Association (JAMIA) suggests we may have to add e-prescribing tools to the list.
When investigators from Harvard and Massachusetts General Hospital reviewed more than 3,800 computer-generated prescriptions for outpatients, they found that 11.7% contained errors, and 35% of those mistakes had the potential to cause real harm--what they refer to as adverse drug reactions (ADRs). Although the research didn't directly compare the electronic errors to those occurring in a handwritten system, the report said: "Our results in terms of error frequency with electronic prescriptions are consistent with outpatient handwritten and electronic error rates that have been reported in the literature."
That statement provides a clue to the investigation's most serious shortcoming. The researchers didn't do a direct case-by-case comparison of electronic and handwritten prescriptions. It's not enough to say that the error rates for handwritten scripts--as "reported in the literature"--were about the same. There's no way to know if the circumstances in these older studies match those in the new Harvard investigation.
If, for instance, the types of drugs ordered or the complexity of the dosing regimens differed in studies reported in the literature, when compared to the Harvard report, an accurate comparison would be impossible.
And if, after reading the latest JAMIA report, you're inclined to re-evaluate e-prescribing tools, keep in mind that its data has to be looked at in the context of all the research that's out there. A review of 25 studies concluded recently that e-prescribing systems do, in fact, reduce the number of medication errors, including those likely to produce an adverse reaction in patients.
Despite the new study's weaknesses, there's still a good deal that IT managers can take away from it. The researchers looked at more than 12 e-prescribing products, and the error rates varied widely among them, from about 5% to almost 38%. The take-home message is clear: Choose your vendor wisely. (Unfortunately, the products weren't mentioned by name in the report.)
Of course, we can't rule out the possibility that this wide range in error rates was due in part to user error, which only emphasizes the importance of making certain clinicians are thoroughly trained on these systems.
The Harvard team outlines several steps to reduce the likelihood of e-prescribing errors.
Forcing functions. One of the most common medication errors that occurs in electronic systems is data omission. If your system forces clinicians to fill in the name of the drug, dosage, and specific indication, and it doesn't allow misleading abbreviations, you're less likely to find such mistakes. In the Harvard study, it was estimated that almost 72% of the medication errors and 63% of the potential ADRs would have been eliminated with forced functions.
Specific decision support, including a feature that automatically checks maximum dosing for various patient populations, would also lower the error rates.
Calculators can help eliminate dispensing errors by calculating the correct dose and duration of treatment based on instructions from the physician, rather than relying on a second person to input the same data elsewhere in the system.
In the end, it's the "preponderance of evidence" that wins the day. A mature e-prescribing system with the right functions, coupled with adequate clinician training, will generate fewer errors than paper and ink.
In the new, all-digital InformationWeek Healthcare: iPads are leading a new wave of devices into the exam room. Are security, tech support, and infection control up to the task? Download it now. (Free registration required.)
InformationWeek Tech Digest, Nov. 10, 2014Just 30% of respondents to our new survey say their companies are very or extremely effective at identifying critical data and analyzing it to make decisions, down from 42% in 2013. What gives?